Early life exposures to crime and criminal justice systems differentiate life course experiences of cohorts and have important implications on mortality and health. Applying the age-period-cohort-interaction model developed by Luo and Hodge (2018; Luo 2015), we analyze the U.S. official arrest data from 1960 to 2014 and estimate two types of cohort effects: inter-cohort variation and intra-cohort life course dynamics. Our findings suggest that a portion of crime trend fluctuations is uniquely attributed to cohort replacement. More importantly, our analysis of intra-cohort life course dynamics—an important dimension of cohort effects neglected in the literature—reveals that some cohorts’ higher than expected arrest rates are mainly driven by arrests at young ages. This finding supports the argument of survivor effects in mortality research and casts doubts to the existing literature that assumes cohort effects to be constant across life course.
Presented in Session 11. Health & Mortality 2